start with getting to feature parity with the centro/treo. Too many things missing to list. Palm knows what they are.
then get voice features working. I don't need them, but most people now demand them.
I had a friend buy a droid and really hate it. He asked me about the Pre cuz he knows I like mine (but he isn't a hacker).
I asked him what things he DID like that he would need if he returned his droid for something else. Almost everything he mentioned is not available (voice search, voice dial, etc) or only available with home brew. No way he was going to try home brew (and darn near ANYONE can make it work).
he was a long time palm user, and his list of must-haves were all missing or preware. He decided to keep his droid because his centro died and he decided to suffer with android rather than give up features. I couldn't argue with that.
I am cointing on palm to be at feature parity with centro/palmOS by the time they ship webOS 2.0, which I predict will be in q1 of next year. If not, pre locks themselves out of the non-technical core treo/blackberry market that currently represents the largest portion of the smartphone market.
ps, talked to a software exec last week who works for a BIG software company that builds cross platform business software. He told me the industry rule of thumb is that they will support anything with 9% to %10 market share. They have a team ready to do webOS but it is only 4-5%, so they are hoping HP makes a play for market share to get there soon. If they sold sll the unsold inventory and didn't have such a high return ratio, they WOULD be at 10% btw.
it isn't the cost of the dev that matters, it's the commitment to 7 years of dev and support expected by enterprise customers even if the platform dies.
HP gets that, and wouldn't have bought Palm if the didn't have plans to back it with 3, 5, and 7 year plans.
i'm expecting great stuff, and i'm expecting the stuff that wll make them a player, not just a faster cpu.